Sinlaku’s forecast Model Output from 00Z, 10 Sept. 2008
Included below is an IR image of Sinlaku with some recent model forecasts (courtesy of FSU):
IR Image and model tracks for Sinlaku. Valid 13Z on the 10th of September, 2008.
The model tracks are in reasonable agreement for the short term: Sinlaku drifting slowly northward over the next few days. Notice how the JTWC track is to the west of many of the model tracks. The long term is a different story.
We’re going to take a closer look at a model that isn’t on this output, the ECMWF. It’s been fairly consistent and the primary model that the T-PARC group are following. Attached, once again is the analyzed AMV 200 hPa flow and the corresponding ECWMF 12-hour forecast.
Comparison between analyzed AMVs (valid at 14Z, September 10th) and ECMWF 12-hour forecast (Valid at 12Z September 10th) of 200 hPa streamlines.
The ECMWF’s 850 Vorticity forecast is shown here. The main track is similar to what it has been: a slow crawl to the north and then a sharp jaunt to the northeast.
But, compare the 00Z run to the previous day’s 12Z forecast. The two forecasts are nearly identical until 12Z on the 14th of September. The newer run keep Sinlaku much closer to the Chinese coast and generally keeps it farther to the west. The landfall time for both forecasts is fairly similar: 00Z on the 17th. The 12Z run brings Sinlaku into central Japan while the 00Z forecast on the 10th brings it into Japan’s southwest coast. Is this a trend?
The differences in the steering flow are a bit tricky to track down, because the storm itself is affecting it’s steering. It seems to be related to it’s position relative to a ridge that is developing to its east. This is worth watching. If the track trends westward, it may pose more of a threat to Taiwan than previously forecast.
The UK Met Office model does exactly that, as shown here, bringing Sinlaku straight across Taiwan by 12Z on the 14th and then into China.
The 00Z GFS model is fairly consistent with the Met Office. The 06Z model seems to be more to the east of the 00Z run (see figure above).
The NOGAPS forecast almost seems a hybrid of the two, bringing Sinlaku into Taiwan and then eastward into Japan.
These slow moving systems are notoriously difficult to predict.